A critical synthesis of remote sensing and machine learning approaches for climate hazard impact on crop yield

This review critically assesses the application of machine learning (ML) algorithms and remote sensing (RS) products in detecting and predicting climate hazards, as well as their impacts on crop yields. Using the PRISMA approach, it examines 177 studies on climate hazards and 197 on RS–ML applicatio...

Descripción completa

Detalles Bibliográficos
Autores principales: Obahoundje, Salomon, Tilahun, Seifu A., Zemadim, Birhanu, Schmitter, Petra
Formato: Journal Article
Lenguaje:Inglés
Publicado: IOP Publishing 2025
Materias:
Acceso en línea:https://hdl.handle.net/10568/177349
_version_ 1855520934319882240
author Obahoundje, Salomon
Tilahun, Seifu A.
Zemadim, Birhanu
Schmitter, Petra
author_browse Obahoundje, Salomon
Schmitter, Petra
Tilahun, Seifu A.
Zemadim, Birhanu
author_facet Obahoundje, Salomon
Tilahun, Seifu A.
Zemadim, Birhanu
Schmitter, Petra
author_sort Obahoundje, Salomon
collection Repository of Agricultural Research Outputs (CGSpace)
description This review critically assesses the application of machine learning (ML) algorithms and remote sensing (RS) products in detecting and predicting climate hazards, as well as their impacts on crop yields. Using the PRISMA approach, it examines 177 studies on climate hazards and 197 on RS–ML applications in crop yield modeling. Research is most concentrated in Asia, followed by Africa and the Americas, with agricultural drought emerging as the most frequently studied hazard. Statistical approaches, such as the coefficient of variation, remain the dominant methods for analyzing climate variability. For drought detection, Random Forest (RF) was the most used ML algorithm (17%), followed by Support Vector Machines (SVM, 11%), Artificial Neural Networks (ANN, 8%), Adaptive Neuro-Fuzzy Inference System (ANFIS, 5%), and Extreme Gradient Boosting (XGBoost, 5%). For drought impacts on crop productivity, RF dominated (39%) followed by Least Absolute Shrinkage and Selection Operator (LASSO, 11%), while for climate variability impacts, RF (21%) led alongside SVM (10%), ANN (9%), Long Short-Term Memory (LSTM) (8%), Multiple Linear Regression (MLR) (8%), and Convolutional Neural Network (CNN) (7%). Asia leads in the integration of advanced ML/DL techniques. In contrast, due to infrastructure and data limitations, Africa predominantly employs simpler and more interpretable models. RS products, namely MODIS, TRMM, CHIRPS, and ERA5, are widely used due to their accessibility. However, their limited spatial resolution restricts their effectiveness. The research gaps include a limited investigation at the sub-national level, insufficient ground-truth validation, and inadequate monitoring of complex, compounding hazards like drought–flood–heatwave interactions. Moreover, the research remains skewed toward economically dominant crops (maize, cotton, and soybeans, neglecting marginal crops (cocoa, cashew, cassava, plantain, and coffee) critical to food-insecure regions. The review recommends hybrid modeling frameworks integrating process-based and data-driven methods, broader spatial and crop coverage, standardized protocols, and real-time, microclimate-aware monitoring systems, for improving model reliability and applicability in underrepresented, data-scarce regions such as sub-Saharan Africa, thereby strengthening climate-resilient agriculture and global food security.
format Journal Article
id CGSpace177349
institution CGIAR Consortium
language Inglés
publishDate 2025
publishDateRange 2025
publishDateSort 2025
publisher IOP Publishing
publisherStr IOP Publishing
record_format dspace
spelling CGSpace1773492025-10-27T04:48:35Z A critical synthesis of remote sensing and machine learning approaches for climate hazard impact on crop yield Obahoundje, Salomon Tilahun, Seifu A. Zemadim, Birhanu Schmitter, Petra climate change impacts weather hazards crop yield remote sensing weather hazards This review critically assesses the application of machine learning (ML) algorithms and remote sensing (RS) products in detecting and predicting climate hazards, as well as their impacts on crop yields. Using the PRISMA approach, it examines 177 studies on climate hazards and 197 on RS–ML applications in crop yield modeling. Research is most concentrated in Asia, followed by Africa and the Americas, with agricultural drought emerging as the most frequently studied hazard. Statistical approaches, such as the coefficient of variation, remain the dominant methods for analyzing climate variability. For drought detection, Random Forest (RF) was the most used ML algorithm (17%), followed by Support Vector Machines (SVM, 11%), Artificial Neural Networks (ANN, 8%), Adaptive Neuro-Fuzzy Inference System (ANFIS, 5%), and Extreme Gradient Boosting (XGBoost, 5%). For drought impacts on crop productivity, RF dominated (39%) followed by Least Absolute Shrinkage and Selection Operator (LASSO, 11%), while for climate variability impacts, RF (21%) led alongside SVM (10%), ANN (9%), Long Short-Term Memory (LSTM) (8%), Multiple Linear Regression (MLR) (8%), and Convolutional Neural Network (CNN) (7%). Asia leads in the integration of advanced ML/DL techniques. In contrast, due to infrastructure and data limitations, Africa predominantly employs simpler and more interpretable models. RS products, namely MODIS, TRMM, CHIRPS, and ERA5, are widely used due to their accessibility. However, their limited spatial resolution restricts their effectiveness. The research gaps include a limited investigation at the sub-national level, insufficient ground-truth validation, and inadequate monitoring of complex, compounding hazards like drought–flood–heatwave interactions. Moreover, the research remains skewed toward economically dominant crops (maize, cotton, and soybeans, neglecting marginal crops (cocoa, cashew, cassava, plantain, and coffee) critical to food-insecure regions. The review recommends hybrid modeling frameworks integrating process-based and data-driven methods, broader spatial and crop coverage, standardized protocols, and real-time, microclimate-aware monitoring systems, for improving model reliability and applicability in underrepresented, data-scarce regions such as sub-Saharan Africa, thereby strengthening climate-resilient agriculture and global food security. 2025-10-23 2025-10-27T04:42:09Z 2025-10-27T04:42:09Z Journal Article https://hdl.handle.net/10568/177349 en Open Access IOP Publishing Obahoundje, S.; Tilahun, S. A.; Zemadim, B.; Schmitter, P. 2025. A critical synthesis of remote sensing and machine learning approaches for climate hazard impact on crop yield. Environmental Research Communications, 7(10):102001. doi: https://doi.org/10.1088/2515-7620/ae1099
spellingShingle climate change impacts
weather hazards
crop yield
remote sensing
weather hazards
Obahoundje, Salomon
Tilahun, Seifu A.
Zemadim, Birhanu
Schmitter, Petra
A critical synthesis of remote sensing and machine learning approaches for climate hazard impact on crop yield
title A critical synthesis of remote sensing and machine learning approaches for climate hazard impact on crop yield
title_full A critical synthesis of remote sensing and machine learning approaches for climate hazard impact on crop yield
title_fullStr A critical synthesis of remote sensing and machine learning approaches for climate hazard impact on crop yield
title_full_unstemmed A critical synthesis of remote sensing and machine learning approaches for climate hazard impact on crop yield
title_short A critical synthesis of remote sensing and machine learning approaches for climate hazard impact on crop yield
title_sort critical synthesis of remote sensing and machine learning approaches for climate hazard impact on crop yield
topic climate change impacts
weather hazards
crop yield
remote sensing
weather hazards
url https://hdl.handle.net/10568/177349
work_keys_str_mv AT obahoundjesalomon acriticalsynthesisofremotesensingandmachinelearningapproachesforclimatehazardimpactoncropyield
AT tilahunseifua acriticalsynthesisofremotesensingandmachinelearningapproachesforclimatehazardimpactoncropyield
AT zemadimbirhanu acriticalsynthesisofremotesensingandmachinelearningapproachesforclimatehazardimpactoncropyield
AT schmitterpetra acriticalsynthesisofremotesensingandmachinelearningapproachesforclimatehazardimpactoncropyield
AT obahoundjesalomon criticalsynthesisofremotesensingandmachinelearningapproachesforclimatehazardimpactoncropyield
AT tilahunseifua criticalsynthesisofremotesensingandmachinelearningapproachesforclimatehazardimpactoncropyield
AT zemadimbirhanu criticalsynthesisofremotesensingandmachinelearningapproachesforclimatehazardimpactoncropyield
AT schmitterpetra criticalsynthesisofremotesensingandmachinelearningapproachesforclimatehazardimpactoncropyield